The purpose of this proposal is to continue and extend the strong core program in biometrical (biostatistical) phases of cancer epidemiology currently going on at RPMI. The three main objectives would be: (1) Continued development of biostatistical technology which would facilitate each of the 9 basic steps in the process of data- utilization. While continuing to develop computer-assisted systems to reduce drudgery and speed access, the emphasis would gradually shift to the evolution of scientific strategies and tactics which would exploit the full potential of the new technology. Once the clerical bottlenecks in the 9 step process are cleared away, the limiting factor often becomes the ability of the epidemiological analyst to cope with large numbers of case histories (10000 plus) and many different variables (100 plus) with many categories per variables (100 plus).(2) The acid test of biostatistical technology is that it actually lead to preventive or therapeutic measures for cancer control. Hence the value of the new technology will be tested by using it to produce new findings in cancer epidemiology and new approaches to cancer control. (3) To improve the dissemination of the new technology. In addition to written or oral presentations, new people would be brought in for periods of a year or two to learn by doing. For the more sophisticated technology--such as deep mathematical models--this is the only way to learn.